Relaxing the Rule of Ten Events per Variable in Logistic and Cox Regression
The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
The rule of thumb that logistic and Cox models should be used with a minimum of 10 outcome events per predictor variable (EPV), based on two simulation studies, may be too conservative. The authors conducted a large simulation study of other influences on confidence interval coverage, type I error, relative bias, and other model performance measures. They found a range of circumstances in which coverage and bias were within acceptable levels despite less than 10 EPV, as well as other factors that were as influential as or more influential than EPV. They conclude that this rule can be relaxed, in particular for sensitivity analyses undertaken to demonstrate adequate control of confounding.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Following the generally accepted rule of thumb that 5–10 events are required per covariate in logit or Cox regression models (Peduzzi et al., 1995; Peduzzi et al., 1996; Vittinghoff & McCulloch, 2007), we assessed whether or not each paper could have supported fully saturated regression models with product terms between all of the psychosocial problems considered.”
From Syndemics of psychosocial problems and HIV risk: A systematic review of empirical tests of the disease interaction concept (2015) · cites this paper as Vittinghoff & McCulloch, 2007
In the citing paper
“Because so few students used psychotherapy, we limited the number of explanatory variables to the number of observed events divided by five (14).”
From Mental health treatment seeking among veteran and civilian community college students (2017) · cites this paper as [14]
In the citing paper
“This sample size was chosen taking into account approximate sample size estimation criteria [38,39]. We worked with the criteria that 10 subjects would be needed per parameter included in a statistical model used to adjust for confounding.”
From Relationships, love and sexuality: what the Filipino teens think and feel (2009) · cites this paper as [38]
In the citing paper
“As Vittinghoff and McCulloch 27 recommended, the 10 outcome events per predictor variable rule was relaxed.”
From Mental health services for parents who lost a child to cancer: if we build them, Will they come? (2015) · cites this paper as [27]
In the citing paper
“The requirements regarding events per variable were fulfilled owing to a considerable number of injuries in the present dataset.34”
From How much running is too much? Identifying high-risk running sessions in a 5200-person cohort study (2025) · cites this paper as [34]
Citations
Cites
A simulation study of the number of events per variable in logistic regression analysis
Peduzzi, P., Concato, J., Kemper, E., Holford, T. R., & Feinstein, A. R. (1996) · Journal of Clinical Epidemiology · doi:10.1016/s0895-4356(96)00236-3 · Cited by 9,655 · cited 1×
“(2, 3) more closely, we also examined models with all binary predictors.”
BackgroundModeling and variable selection in epidemiologic analysis
Greenland, S. (1989) · American Journal of Public Health · doi:10.2105/ajph.79.3.340 · Cited by 2,208
Peduzzi, P., Concato, J., Feinstein, A. R., & Holford, T. R. (1995) · Journal of Clinical Epidemiology · doi:10.1016/0895-4356(95)00048-8 · Cited by 2,154 · cited 2×
“(2, 3) more closely, we also examined models with all binary predictors.”
BackgroundResultThe Abuse of Power: The Pervasive Fallacy of Power Calculations for Data Analysis
Hoenig, J. M., & Heisey, D. M. (2001) · The American Statistician · doi:10.1198/000313001300339897 · Cited by 2,046
Sample-Size Formula for the Proportional-Hazards Regression Model
Schoenfeld, D. A. (1983) · Biometrics · doi:10.2307/2531021 · Cited by 1,096
Concato, J., Peduzzi, P., Holford, T. R., & Feinstein, A. R. (1995) · Journal of Clinical Epidemiology · doi:10.1016/0895-4356(95)00510-2 · Cited by 797 · cited 1×
“The rule of thumb that logistic and Cox models should be used with a minimum of 10 events per predictor variable (EPV) is based on two simulation studies (1–3).”
MethodSection 4 Long-term survival results in medically and surgically randomized patients
Detre, K. M., Peduzzi, P. N., Scott, S. M., & Davies, B. (1985) · Progress in Cardiovascular Diseases · doi:10.1016/0033-0620(85)90019-2 · Cited by 24
Section 2 Design and baseline characteristics
Peduzzi, P. N., Detre, K. M., & Gage, A. A. (1985) · Progress in Cardiovascular Diseases · doi:10.1016/0033-0620(85)90017-9 · Cited by 15
Harrell, F., Jr. (1996) · doi:10.1002/(sici)1097-0258(19960229)15:4<361::aid-sim168>3.0.co;2-4
Cited by
Items 501–600 of 1,018
Berruezo, A., Penela, D., Jáuregui, B., Soto‐Iglesias, D., Aguinaga, L., Ordóñez, A., Fernández‐Armenta, J., Martínez, M., Tercedor, L., Bisbal, F., et al. (2019) · EP Europace · doi:10.1093/europace/euz027 · Cited by 51
Čukić, M., López, V., & Pavón, J. (2020) · Journal of Medical Internet Research · doi:10.2196/19548 · Cited by 45
“A minimum rate of 10 cases per predictor is common [64], although it is not a universal recommendation [67].”
BackgroundFactors associated with delirium and cognitive decline following hip fracture surgery
Uzoigwe, C. E., O’Leary, L., Nduka, J. C., Sharma, D., Melling, D., Simmons, D., & Barton, S. (2020) · The Bone & Joint Journal · doi:10.1302/0301-620x.102b12.bjj-2019-1537.r3 · Cited by 50
Baeza-Delgado, C., Alberich, L. C., Sierra, J. M. C., Veiga-Canuto, D., Heras, B. M. D. L., Raza, B., & Martí‐Bonmatí, L. (2022) · European Radiology Experimental · doi:10.1186/s41747-022-00276-y · Cited by 55
“In order to compare the sample size obtained with other accepted methodologies, sample sizes were also calculated following the 10 EPP “rule of thumb” [7–9] and the 5 EPP estimation [12].”
MethodTortorella, G. L., Fogliatto, F. S., Kumar, M., González, V. A., & Pepper, M. (2022) · Journal of Manufacturing Technology Management · doi:10.1108/jmtm-04-2022-0173 · Cited by 59
Outcomes of Patients With Atrial Fibrillation Following Thrombectomy for Stroke
Kobeissi, H., Ghozy, S. A., Seymour, T., Gupta, R., Bilgin, C., Kadirvel, R., Rabinstein, A. A., & Kallmes, D. F. (2023) · JAMA Network Open · doi:10.1001/jamanetworkopen.2022.49993 · Cited by 45
Pollak, M., Leroy, S., Röhr, V., Brown, E. N., Spies, C., & Koch, S. (2024) · Anesthesiology · doi:10.1097/aln.0000000000004929 · Cited by 48
Hua, Y., Stead, T. S., George, A., & Ganti, L. (2025) · Academic Medicine & Surgery · doi:10.62186/001c.131964 · Cited by 69
Psychosocial impact of involvement in the Special Olympics
Crawford, C., Burns, J., & Fernie, B. A. (2015) · Research in Developmental Disabilities · doi:10.1016/j.ridd.2015.07.009 · Cited by 65
Messika, J., Sztrymf, B., Bertrand, F., Billard‐Pomares, T., Barnaud, G., Branger, C., Dreyfuss, D., & Ricard, J.-D. (2014) · Journal of Critical Care · doi:10.1016/j.jcrc.2014.05.021 · Cited by 45
Chao, T., Tsao, H.-M., Kibos, A., Lin, Y., Lin, W., Chang, S., Lo, L., Hu, Y., Tuan, T., Suenari, K., et al. (2012) · Heart Rhythm · doi:10.1016/j.hrthm.2012.06.039 · Cited by 45
Linden, B. C., Bairdain, S., Zurakowski, D., Shamberger, R. C., & Lillehei, C. W. (2013) · Journal of Pediatric Surgery · doi:10.1016/j.jpedsurg.2012.08.031 · Cited by 45
Hunchak, C., Teklu, S., Meshkat, N., Meaney, C. A., & Ritchie, L. M. P. (2015) · BMC Research Notes · doi:10.1186/s13104-015-1592-z · Cited by 45
“significant mortality reduction from simple interventions including ED care re-organization and may contain some important lessons to adapt to similarly resourced adult ED settings [12].”
BackgroundMethodHübscher, M., Moloney, N. A., Rebbeck, T. J., Traeger, A. C., & Refshauge, K. M. (2013) · Clinical Journal of Pain · doi:10.1097/ajp.0000000000000045 · Cited by 50
Mulders, A., Fick, I. W. F., Bor, H. H. J., Verhey, F. R. J., Zuidema, S. U., & Koopmans, R. T. C. M. (2016) · Journal of the American Medical Directors Association · doi:10.1016/j.jamda.2016.01.002 · Cited by 52
Brainin, P., Hoffmann, S., Fritz‐Hansen, T., Olsen, F. J., Jensen, J. S., & Biering‐Sørensen, T. (2018) · Journal of the American Society of Echocardiography · doi:10.1016/j.echo.2018.05.007 · Cited by 49
Roth, J. A., Juchler, F., Dangel, M., Eckstein, F. S., Battegay, M., & Widmer, A. F. (2018) · Clinical Infectious Diseases · doi:10.1093/cid/ciy879 · Cited by 52
Cognitive Impairment is Associated with Mortality in Hemodialysis Patients
Angermann, S., Schier, J., Baumann, M., Steubl, D., Hauser, C., Lorenz, G., Günthner, R., Braunisch, M. C., Kemmner, S., Satanovskij, R. M., et al. (2018) · Journal of Alzheimer s Disease · doi:10.3233/jad-180767 · Cited by 48
“As the number of predictors that can be included in the multivariate Cox regression model is limited by the number of events [27, 28], 7 to 10 predictor variables could be included in the multivariate Cox regression model, given the number of 69 events in this study cohort [13].”
Peng, Q., Zhang, Y.-M., Liang, L., Liu, X., Ye, L.-F., Yang, H.-B., Zhang, L., Shu, X.-M., Lu, X., & Wang, G.-C. (2019) · Clinical & Experimental Immunology · doi:10.1111/cei.13404 · Cited by 45
Allin, J. R., Hunt, J. E., Talling, P. J., Clare, M. A., Pope, E., & Masson, D. G. (2015) · Marine Geology · doi:10.1016/j.margeo.2015.11.005 · Cited by 50
Incidence of thromboembolic events in asymptomatic carriers of IgA anti ß2 glycoprotein-I antibodies
Tortosa, C., Cabrera-Marante, O., Serrano, M., Martínez-Flores, J. A., Pérez, D., Lora, D., Morillas, L., Paz-Artal, E., Morales, J. M., Pleguezuelo, D., et al. (2017) · PLoS ONE · doi:10.1371/journal.pone.0178889 · Cited by 44
Ebert, M. A., Gulliford, S., Acosta, O., de Crevoisier, R., McNutt, T., Heemsbergen, W. D., Witte, M. G., Palma, G., Rancati, T., & Fiorino, C. (2021) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ac0681 · Cited by 50
A Meta-Regression of Racial Disparities in Wellbeing Outcomes During and After Foster Care
Kennedy, R. S., Potter, M. H., & Font, S. A. (2023) · Trauma, Violence, & Abuse · doi:10.1177/15248380221111481 · Cited by 44
Covariate adjustment had similar benefits in small and large randomized controlled trials
Thompson, D. D., Lingsma, H. F., Whiteley, W. N., Murray, G. D., & Steyerberg, E. W. (2015) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2014.11.001 · Cited by 44
“Some authors recommend a minimum of 10 events per variable (EPV) (24), while an EPV of 5 may be sufficient in the context of adjusting for confounders (25).”
BackgroundRoelen, C. A. M., van Hoffen, M. F. A., Groothoff, J. W., de Bruin, J., Schaufeli, W. B., & van Rhenen, W. (2014) · International Archives of Occupational and Environmental Health · doi:10.1007/s00420-014-0981-2 · Cited by 53
“Although there were few mental, musculoskeletal, and other somatic LTSA events, the number of events per variable was sufficient for stable predictions (Vittinghoff and McCulloch 2007), because confounder analysis restricted the number of independent variables in the logistic regression models.”
BackgroundFeng, Y., Maurel, N., Wang, Z., Lei, N., Yu, F., & van Kleunen, M. (2016) · Global Ecology and Biogeography · doi:10.1111/geb.12497 · Cited by 45
“…had sufficient ability to produce reliable parameter estimates, we checked whether the num ber of events per explanatory variable was greater than five (Vittinghoff & McCulloch, 2007), and whether the variance inflation factor (indicating multicollinearity) was less than four (Zuur et al., 2009).”
MethodLuna, I. E., Kehlet, H., Petersen, M. A., & Aasvang, E. K. (2017) · Acta Anaesthesiologica Scandinavica · doi:10.1111/aas.12899 · Cited by 49
FULLER, J. T., Thewlis, D., Buckley, J. D., Brown, N. A. T., Hamill, J., & Tsiros, M. D. (2017) · The American Journal of Sports Medicine · doi:10.1177/0363546516682497 · Cited by 55
“Two separate models were used for analysis to maximize the number of injury events per model factor.(40) Model 1 included shoe as a categorical factor, weekly training distance as a continuous predictor, and shoe 3 weekly training distance interaction.”
BackgroundMethodRemiro‐Azócar, A., Heath, A., & Baio, G. (2022) · Research Synthesis Methods · doi:10.1002/jrsm.1565 · Cited by 52
“The difference in results is likely due to logistic regression being more prone to small-sample bias than Cox regression.(83) As for precision, the G-computation approaches have reduced variability compared to MAIC.”
ResultBoosting the impact of knowledge management on innovation performance through industry 4.0 adoption
Tortorella, G. L., Prashar, A., Vassolo, R. S., Vergara, A., Filho, M. G., & Samson, D. (2022) · Knowledge Management Research & Practice · doi:10.1080/14778238.2022.2108737 · Cited by 54
“However, Concato et al. (1995), Peduzzi et al. (1995), and Vittinghoff and McCulloch (2007) suggested that there should be at least 10 observations per variable in a regression analysis.”
BackgroundIndividual Pathways to Life Satisfaction: The Significance of Traits and Motives
Hofer, J., Busch, H., & Kießling, F. (2008) · Journal of Happiness Studies · doi:10.1007/s10902-007-9086-x · Cited by 46
“Although according to common statistical guidelines the present sample size is sufficient for the analyses at hand (Tabachnik and Fidell 1996; Vittinghoff and McCulloch 2007 ), it would certainly be informative to replicate the current analyses with a bigger sample.”
Batelaan, N. M., ten Have, M., van Balkom, A. J. L. M., Tuithof, M., & de Graaf, R. (2014) · Journal of Anxiety Disorders · doi:10.1016/j.janxdis.2013.12.003 · Cited by 54
Predicting patients with concurrent noncontiguous spinal epidural abscess lesions
Ju, K. L., Kim, S. D., Melikian, R., Bono, C. M., & Harris, M. B. (2014) · The Spine Journal · doi:10.1016/j.spinee.2014.06.008 · Cited by 43
Cohort studies of aetiology and prognosis: they’re different
Herbert, R. (2014) · Journal of physiotherapy · doi:10.1016/j.jphys.2014.07.005 · Cited by 55
Bosch, W., Heckman, M. G., Pungpapong, S., Diehl, N. N., Shalev, J. A., & Hellinger, W. C. (2012) · Transplantation · doi:10.1097/tp.0b013e3182472876 · Cited by 43
“With that variable included in all models, we then adjusted for variables showing the strongest single variable associations with the given endpoint, allowing for no more than one variable in the model for every 10 patients who experienced the given endpoint (26).”
MethodSchau, T., Seifert, M., Meyhöfer, J., Neuß, M., & Butter, C. (2011) · EP Europace · doi:10.1093/europace/eur153 · Cited by 43
Physical and Emotional Predictors of Depression After Radical Prostatectomy
Weber, B. A., Roberts, B. L., Mills, T. L., Chumbler, N. R., & Algood, C. B. (2008) · American Journal of Men s Health · doi:10.1177/1557988307312222 · Cited by 43
“The general rule of thumb is that 10 participants per independent variable are needed for adequate power in logistic regression analysis (Hosmer & Lemeshow, 2000; Peduzzi, Concato, Kemper, Holford, & Feinstein, 1996; Vittinghoff & McCulloch, 2006).”
BackgroundMethodPanman, C. M. C. R., Wiegersma, M., Kollen, B. J., Burger, H., Berger, M. Y., & Dekker, J. H. (2016) · International Urogynecology Journal · doi:10.1007/s00192-016-3107-4 · Cited by 53
“The rule of thumb that logistic regression models should be used with a minimum of 10 events per predictor variable (EPV) is based on very few simulation studies in which only the numbers of events was varied [23].”
BackgroundMethodMoxon, J. V., Jones, R. E., Wong, G., Weir, J. M., Mellett, N. A., Kingwell, B. A., Meikle, P. J., & Golledge, J. (2017) · Atherosclerosis · doi:10.1016/j.atherosclerosis.2017.06.925 · Cited by 49
“MonteCarlo simulations suggest that 10 ou tc me events are required for each 319 covariate entered into regression models, although there are suggestions that this may be 320 overly stringent [38].”
BackgroundTime since prior wildfire affects subsequent fire containment in black spruce
Beverly, J. L. (2017) · International Journal of Wildland Fire · doi:10.1071/wf17051 · Cited by 48
Devulapalli, K. K., Fidelman, N. A., Soulen, M. C., Miller, M. D., Johnson, M. S., Addo, E. A., El-Haddad, G. E., Nutting, C. W., Morrison, J., Farsad, K., et al. (2018) · Radiology · doi:10.1148/radiol.2018170962 · Cited by 43
Do Stem Design and Surgical Approach Influence Early Aseptic Loosening in Cementless THA?
Janssen, L., Wijnands, K. A. P., Janssen, D. W., Janssen, M. W. H. E., & Morrenhof, J. W. (2018) · Clinical Orthopaedics and Related Research · doi:10.1007/s11999.0000000000000208 · Cited by 44
de Jong, Y., Ramspek, C. L., van der Endt, V. H. W., Rookmaaker, M. B., Blankestijn, P. J., Vernooij, R. W. M., Verhaar, M. C., Bos, W. J. W., Dekker, F. W., Ocak, G., et al. (2020) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2020.03.015 · Cited by 49
Sharif‐Askari, F. S., Sulaiman, S. A. S., Sharif‐Askari, N. S., & Hussain, A. A. S. (2014) · PLoS ONE · doi:10.1371/journal.pone.0095991 · Cited by 52
“To prevent model over fitting, the maximum number of variables entered in the multivariate regression models was one variable for every eight ADR events [26].”
Epidemiology, Prognostic Factors, and Treatment of Malignant Submandibular Gland Tumors
Lee, R. J., Tan, A. P., Tong, E. L., Satyadev, N., & Christensen, R. E. (2015) · JAMA Otolaryngology–Head & Neck Surgery · doi:10.1001/jamaoto.2015.1745 · Cited by 42
Martelli, G., Boracchi, P., Orenti, A., Lozza, L., Maugeri, I., Vetrella, G., & Agresti, R. (2014) · European Journal of Surgical Oncology · doi:10.1016/j.ejso.2014.03.029 · Cited by 42
A New Method for Evaluating Trauma Centre Outcome Performance: TRAM-Adjusted Mortality Estimates
Moore, L., Hanley, J. A., Turgeon, A. F., Lavoie, A., & Eric, B. (2010) · Annals of Surgery · doi:10.1097/sla.0b013e3181d97589 · Cited by 46
“To address the problem of assessing performance in lowvolume hospitals, we used the guideline of a minimum of 5 events per variable to flag hospitals that may not provide informative mortality estimates.(38) To avoid penalizing high-volume hospitals with low mortality, the minimum number of patients required was estim…”
BackgroundMethodKemp, A., Preen, D. B., Morlet, N., Clark, A., McAllister, I. L., Briffa, T. G., Sanfilippo, F. M., Ng, J. Q., McKnight, C. M., Reynolds, W., et al. (2013) · Retina · doi:10.1097/iae.0b013e318276e07b · Cited by 43
Grant, J. A., Bedi, A., Kurz, J., Bancroft, R., Gagnier, J. J., & Miller, B. S. (2015) · Sports Health: A Multidisciplinary Approach · doi:10.1177/1941738114540445 · Cited by 42
Impaired thrombolysis: a novel cardiovascular risk factor in end-stage renal disease
Sharma, S., Farrington, K., Kozarski, R., Christopoulos, C., Niespialowska-Steuden, M., Moffat, D., & Gorog, D. A. (2012) · European Heart Journal · doi:10.1093/eurheartj/ehs300 · Cited by 42
“Based on the assumption that we would see around 14% MACE events per year in the ESRD cohort,(34,35) a sample size of 200 was predicted to yield 30 events/year and the final Cox PH model with four covariates would provide 5% significance and 80% power.(36,37) Given 200 patients in the intervention group and arbitrary…”
MethodJuvenile rough-and-tumble play predicts adult sexual behaviour in American mink
Dallaire, J. A., & Mason, G. J. (2016) · Animal Behaviour · doi:10.1016/j.anbehav.2016.10.023 · Cited by 45
PREDIMED Study Investigators, Downer, M. K., Martínez‐González, M. Á., Gea, A., Stampfer, M. J., Wärnberǵ, J., Ruiz‐Canela, M., Salas‐Salvadó, J., Corella, D., Ros, E., et al. (2017) · BMC Cardiovascular Disorders · doi:10.1186/s12872-016-0435-8 · Cited by 44
Neovascular glaucoma after vitrectomy in patients with proliferative diabetic retinopathy
Kwon, J. W., Jee, D., & La, T. Y. (2017) · Medicine · doi:10.1097/md.0000000000006263 · Cited by 46
“…the development of NVG after adjusting for serum glucose levels, hemoglobin A1c levels, blood pressure, preoperative PRP, lens status, and the time from PRP to vitreous hemorrhage, which were all possible confounding factors affecting retinal ischemia (OR, 4.43; CI, 1.28–15.27; P=0.019; Table 3).”
BackgroundInfluentialGoodman, K. E., Lessler, J., Harris, A. D., Milstone, A. M., & Tamma, P. D. (2019) · Infection Control & Hospital Epidemiology · doi:10.1017/ice.2019.17 · Cited by 42
“18,19 In contrast, CART is nonparametric and makes fewer data assumptions, 13 and it can accommodate collinear independent variables.”
BackgroundAnatomic location of primary melanoma: Survival differences and sun exposure
Howard, M. D., Wee, E., Wolfe, R. S., McLean, C., Kelly, J. W., & Pan, Y. (2019) · Journal of the American Academy of Dermatology · doi:10.1016/j.jaad.2019.04.034 · Cited by 42
Weinkle, A. P., Pettit, C. J., Jani, A. H., Keller, J. J., Lu, Y., Malachowski, S. J., Trinidad, J. C., Kaffenberger, B. H., Ergen, E. N., Hughey, L. C., et al. (2019) · Journal of the American Academy of Dermatology · doi:10.1016/j.jaad.2019.05.061 · Cited by 48
Coiro, S., Simonovic, D., Deljanin-Ilić, M., Duarte, K., Carluccio, E., Cattadori, G., Girerd, N., & Ambrosio, G. (2020) · Circulation Heart Failure · doi:10.1161/circheartfailure.119.006769 · Cited by 46
Möhlenkamp, S., Lehmann, N., Greenland, P., Moebus, S., Kälsch, H., Schmermund, A., Dragano, N., Stang, A., Siegrist, J., Mann, K., et al. (2010) · Atherosclerosis · doi:10.1016/j.atherosclerosis.2010.12.014 · Cited by 42
Feline Upper Respiratory Tract Lymphoma
Santagostino, S. F., Mortellaro, C. M., Boracchi, P., Avallone, G., Caniatti, M., Forlani, A., & Roccabianca, P. (2014) · Veterinary Pathology · doi:10.1177/0300985814537529 · Cited by 41
“According to the caseload included in the study, a ratio of 1:5 was considered, as previously suggested.(36) Since it was not possible to include all the above-mentioned variables, only a subset of variables were jointly analyzed in multivariable regression models (mitotic index, age, FeLV status, phenotype, and epith…”
MethodQuinn, C., Uzbeck, M. H., Saleem, I., Cotter, P. E., Ali, J., O’Malley, G., Gilmartin, J. J., & O’Keeffe, S. T. (2011) · Sleep Medicine · doi:10.1016/j.sleep.2010.08.014 · Cited by 48
“96 [9–712], Mann– Whitney U 7344, p = 0.”
BackgroundMethodEarp, B. E., Han, C. H., Floyd, W. E., Rozental, T. D., & BLAZAR, P. E. (2015) · The Journal Of Hand Surgery · doi:10.1016/j.jhsa.2014.12.027 · Cited by 43
Wynants, L., Vergouwe, Y., Van Huffel, S., Timmerman, D., & Van Calster, B. (2018) · Statistical Methods in Medical Research · doi:10.1177/0962280216668555 · Cited by 41
Muscle-strengthening and aerobic activities and mortality among 3+ year cancer survivors in the U.S.
Tarasenko, Y. N., Linder, D. F., & Miller, E. A. (2018) · Cancer Causes & Control · doi:10.1007/s10552-018-1017-0 · Cited by 42
Tan, D. J. A., Sultana, R., Han, N. L. R., Sia, A. T. H., & Sng, B. L. (2018) · BMC Anesthesiology · doi:10.1186/s12871-018-0514-8 · Cited by 45
Impact of post‐procedural minimal stent area on 2‐year clinical outcomes in the SYNTAX II trial
Katagiri, Y., Maria, G. L. D., Kogame, N., Chichareon, P., Takahashi, K., Chang, C. C., Modolo, R., Walsh, S. J., Sabaté, M., Davies, J. E., et al. (2019) · Catheterization and Cardiovascular Interventions · doi:10.1002/ccd.28105 · Cited by 43
“Seventh, the low number of events per predictor may cause overfitting of the multivariate model, although five events per predictor could be acceptable.(31) Eighth, although post-procedural MSA was an independent predictor of TLR in the present analysis, the c-statistic in the ROC analysis was relatively low.”
BackgroundCrawford, B. J., Sourki, R., Khayyam, H., & Milani, A. S. (2021) · Computers in Industry · doi:10.1016/j.compind.2021.103510 · Cited by 42
Surgical factors associated with symptomatic implant removal after patella fracture
Gupta, J., Harkin, E., O’Connor, K., Enobun, B., O’Hara, N. N., & O’Toole, R. V. (2022) · Injury · doi:10.1016/j.injury.2022.03.028 · Cited by 40
Yuan, X., Xu, Q., Du, F., Gao, X., Guo, J., Zhang, J., Wu, Y., Zhou, Z., Yu, Y., & Zhang, Y. (2025) · EClinicalMedicine · doi:10.1016/j.eclinm.2024.103023 · Cited by 58
Oosting, E., Hoogeboom, T. J., Vries, S. A. A.-D., Swets, A., Dronkers, J. J., & van Meeteren, N. L. U. (2015) · Disability and Rehabilitation · doi:10.3109/09638288.2015.1076074 · Cited by 43
“The sample size was based on five events per variable [28,29].”
BackgroundLin, I., Chang, W. P., & Liao, Y.-N. (2013) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2013.02.002 · Cited by 43
Keyser, T. L., & Loftis, D. L. (2015) · New Forests · doi:10.1007/s11056-015-9470-z · Cited by 43
Raman spectroscopic grading of astrocytoma tissues: using soft reference information
Beleites, C., Geiger, K. D., Kirsch, M., Sobottka, S. B., Schackert, G., & Salzer, R. (2011) · Analytical and Bioanalytical Chemistry · doi:10.1007/s00216-011-4985-4 · Cited by 43
“Many spectroscopic studies aiming at medical diagnosis comprise patient numbers 10 to 100 times below the recommended five samples per class and variate [28–31].”
BackgroundZhao, X., Rødland, E. A., Sørlie, T., Vollan, H. K. M., Russnes, H. G., Kristensen, V. N., Lingjærde, O. C., & Børresen‐Dale, A. (2014) · BMC Cancer · doi:10.1186/1471-2407-14-211 · Cited by 44
Price, A. M. H., Wake, M. A., Ukoumunne, O. C., & Hiscock, H. (2012) · Sleep Medicine · doi:10.1016/j.sleep.2012.04.014 · Cited by 45
“The adjusted regression models included the five potential confounders and randomization status (whether the child was assigned to the intervention or control group), except that ‘‘moderate/large’’ child sleep problems and the global child–parent relationship rating analyses were not adjusted, and analysis of child gl…”
MethodLung density associates with survival in alpha 1 antitrypsin deficient patients
Green, C. E., Parr, D. G., Edgar, R. G., Stockley, R. A., & Turner, A. M. (2016) · Respiratory Medicine · doi:10.1016/j.rmed.2016.01.007 · Cited by 40
“We took a conservative statistical approach, consistent with classical teaching to include nomore than one variable per ten events [22], aimed at maximizing power and minimizing bias, although arguments can be presented for taking a broader statistical approach [16].”
MethodResultPalatini, P., Fania, C., Mos, L., Garavelli, G., Mazzer, A., Cozzio, S., Saladini, F., & Casiglia, E. (2016) · International Journal of Cardiology · doi:10.1016/j.ijcard.2016.03.006 · Cited by 48
“More recent results by Vittinghoff and McCulloch have shown that analyses with as few as five events per variable are still statistically valid in samples >1000 subjects especially for significant associations hypothesized a priori (53).”
ResultKasprowicz, M., Burzyńska, M., Melcer, T., & Kübler, A. (2016) · British Journal of Neurosurgery · doi:10.3109/02688697.2016.1161173 · Cited by 47
Associations between osteoporosis and coronary artery disease in postmenopausal women
Lee, S., Cho, J. S., Eun, Y., Song, S.-W., & Moon, K.-W. (2016) · Climacteric · doi:10.1080/13697137.2016.1200550 · Cited by 45
Stein, D. J., Karam, E. G., Shahly, V., Hill, E. D., King, A., Petukhova, M., Atwoli, L., Bromet, E. J., Florescu, S., Haro, J. M., et al. (2016) · BMC Psychiatry · doi:10.1186/s12888-016-0957-8 · Cited by 40
“Even though more recent statistical evaluations show that this rule often can be relaxed [42] and that other considerations (e.”
BackgroundClark, N. P., Douketis, J. D., Hasselblad, V., Schulman, S., Kindzelski, A., & Ortel, T. L. (2017) · American Heart Journal · doi:10.1016/j.ahj.2017.09.015 · Cited by 44
“Event per variable ratios b5 are vulnerable to type I error and bias away from the null hypothesis.(22) Our findings should be interpreted with caution because a low event per variable ratio increases the risk of falsely identifying predictors or overestimating their respective importance.”
Lee, F., Yang, P., Chien, M.-N., Lee, J.-J., Leung, C.-H., & Cheng, S. (2018) · International Journal of Medical Sciences · doi:10.7150/ijms.28498 · Cited by 41
“However, the validity of this thumb rule has been questioned, and the rule of 10 EPV could be relaxed, depending on the research question [35].”
BackgroundGallagher, K. E., Knoll, M. D., Prosperi, C., Baggett, H. C., Brooks, W. A., Feikin, D. R., Hammitt, L. L., Howie, S. R. C., Kotloff, K. L., Levine, O. S., et al. (2019) · Clinical Infectious Diseases · doi:10.1093/cid/ciz350 · Cited by 52
“A total of 1802 HIV-negative cases in the development dataset and 120 deaths allowed us to produce a robust model containing ≤12 variables [14, 15].”
MethodLandes, S. D., Turk, M. A., Damiani, M. R., Proctor, P., & Baier, S. (2021) · JAMA Network Open · doi:10.1001/jamanetworkopen.2021.12862 · Cited by 46
Alvarado‐Bolaños, A., Ayan, D. P., Khaw, A. V., Mai, L. M., Mandzia, J. L., Bogiatzi, C., Mrkobrada, M., Bullrich, M. B., Fleming, L. A., Lippert, C., et al. (2023) · Stroke · doi:10.1161/strokeaha.123.043672 · Cited by 40
González‐Bustamante, B., & Olivares, A. (2016) · Colombia Internacional · doi:10.7440/colombiaint87.2016.04 · Cited by 39
Short parenteral antibiotic treatment for adult septic arthritis after successful drainage
Uçkay, İ., Tovmirzaeva, L., Garbino, J., Rohner, P., Tahintzi, P., Suvà, D., Assal, M., Hoffmeyer, P. J., Bernard, L., & Lew, D. (2012) · International Journal of Infectious Diseases · doi:10.1016/j.ijid.2011.12.019 · Cited by 46
Servais, S., Porcher, R., Xhaard, A., Robin, M., Masson, E., Larghero, J., Ribaud, P., Dhédin, N., Abbes, S., de Fontbrune, F. S., et al. (2013) · Haematologica · doi:10.3324/haematol.2013.089979 · Cited by 39
“001 Female/male 113 (26) 82 (29) 31 (19) Other combinations 329 (74) 196 (71) 133 (81)”
InfluentialVariation in the Diagnosis and Management of Appendicitis at Canadian Pediatric Hospitals
Thompson, G. C., Schuh, S., Gravel, J., Reid, S. M., Fitzpatrick, E. A., Turner, T. W. S., Bhatt, M., Beer, D. L., Blair, G. K., Eccles, R., et al. (2015) · Academic Emergency Medicine · doi:10.1111/acem.12709 · Cited by 45
van der Worp, M. P., de Wijer, A., van Cingel, R., Verbeek, A. L. M., Nijhuis-van der Sanden, M. W. G., & Staal, J. B. (2016) · Journal of Orthopaedic & Sports Physical Therapy · doi:10.2519/jospt.2016.6402 · Cited by 33
Lim, S., Kesavabhotla, K., Cybulski, G. R., Dahdaleh, N. S., & Smith, Z. A. (2016) · Spine · doi:10.1097/brs.0000000000001737 · Cited by 39
“2 and five more occurrences.(27,28) We utilized multivariate logistic regression model to assess the predictive value of each candidate risk factors and their adjusted odds ratios (ORs) were calculated.”
Rickard, C. M., Edwards, M., Spooner, A. J., Mihala, G., Marsh, N. M., Best, J., Wendt, T., Rapchuk, I. L., Gabriel, S., Thomson, B., et al. (2016) · Journal of Critical Care · doi:10.1016/j.jcrc.2016.06.006 · Cited by 39
“Rules of thumb limiting the number of covariates based on the sample size [25] and the number of outcome events [26] were also considered.”
BackgroundResearch on the relationship between urban form and urban smog in China
Liu, Y., Arp, H. P. H., Song, X., & Song, Y. (2016) · Environment and Planning B Urban Analytics and City Science · doi:10.1177/0265813515624687 · Cited by 39
“Peduzzi et al. (1996) recommends not to have more than one independent predictor variable per 10 data points, though Vittinghoff and McCulloch (2007) responded a minimum of 10 outcome events per predictor variable may be too conservative under certain circumstances.”
BackgroundPain-related anxiety as a predictor of early lapse and relapse to cigarette smoking.
LaRowe, L. R., Langdon, K. J., Zvolensky, M. J., Zale, E. L., & Ditre, J. W. (2017) · Experimental and Clinical Psychopharmacology · doi:10.1037/pha0000127 · Cited by 42
“Established procedures for the Cox proportional hazards model indicate that a minimum of 5 events should be included per predictor variable to increase confidence interval coverage, and decrease relative bias and type I error (Vittinghoff & McCulloch, 2007).”
Development of an ICU discharge instrument predicting psychological morbidity: a multinational study
Milton, A., Schandl, A. R., Soliman, I. W., Meijers, K., van den Boogaard, M., Larsson, I., Brorsson, C., Östberg, U., Collet, M. O., Savilampi, J., et al. (2018) · Intensive Care Medicine · doi:10.1007/s00134-018-5467-3 · Cited by 47
Chinese multinationals’ FDI motivations: suggestion for a new theory
Park, B. I., & Roh, T. (2018) · International Journal of Emerging Markets · doi:10.1108/ijoem-03-2017-0104 · Cited by 46
Idiopathic Retroperitoneal Fibrosis: Long-term Risk and Predictors of Relapse
Raffiotta, F., Escoli, R., Quaglini, S., Rognoni, C., Sacchi, L., Binda, V., Messa, P., & Moroni, G. L. (2019) · American Journal of Kidney Diseases · doi:10.1053/j.ajkd.2019.04.020 · Cited by 44
Wong, Y., Cheung, C. Y. Y., Tang, C. S.-M., Hai, J. S. H., Lee, C., Lau, K. K., Au, K., Cheung, B. M. Y., Sham, P. C., Xu, A., et al. (2019) · Cardiovascular Diabetology · doi:10.1186/s12933-019-0974-2 · Cited by 41
Corrected Minute Ventilation Is Associated With Mortality in ARDS Caused by COVID-19
Fusina, F., Albani, F., Bertelli, M., Cavallo, E., Crisci, S., Caserta, R., Nguyen, M., Grazioli, M., Schivalocchi, V., Rosanò, A., et al. (2020) · Respiratory Care · doi:10.4187/respcare.08314 · Cited by 39
“We estimated the need to observe 90 outcomes, therefore 9 variables could be included in the final model.(16) Variables were described with mean 6 SD or median (interquartile range) as appropriate, while factor variables were described as count (percentage).”
Signals in equity-based crowdfunding and risk of failure
Reichenbach, F., & Walther, M. (2021) · Financial Innovation · doi:10.1186/s40854-021-00270-0 · Cited by 55
“In logistic and Cox regressions, the number of events occurring is decisive for the potential number of variables to be tested (Vittinghoff and McCulloch 2007).”
BackgroundMeißner, A.-K., Gutsche, R., Galldiks, N., Köcher, M., Jünger, S. T., Eich, M., Montesinos‐Rongen, M., Brunn, A., Deckert, M., Wendl, C. M., et al. (2021) · Neuro-Oncology · doi:10.1093/neuonc/noab294 · Cited by 45
Metadata
- Authors
- Vittinghoff, E., & McCulloch, C. E.
- Affiliations
- University of California, San Francisco (US)
- Venue
- American Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2007-01-12
- Volume
- 165
- Issue
- 6
- Pages
- 710-718
- ISSN
- 0002-9262 (ISSN-L)
- 1476-6256
- OpenAlex
- W2130373985
- Cited by
- 4,359
- References
- 9
- Citations by year
- 2026: 642 · 2025: 394 · 2024: 313 · 2023: 315 · 2022: 306 · 2021: 351 · 2020: 332 · 2019: 281 · 2018: 282 · 2017: 205 · 2016: 194 · 2015: 171 · 2014: 146 · 2013: 132 · 2012: 94
- FWCI
- 20.12
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- American Journal of Epidemiology (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Keywords
- sample size, logistic regression, Cox proportional hazards model, confidence interval coverage, type I error, relative bias, simulation study, sensitivity analysis, confounding adjustment, small sample size
- MeSH terms
- Data Interpretation, Statistical, Factor Analysis, Statistical, Humans, Predictive Value of Tests, Sensitivity and Specificity, Bias, Confounding Factors, Epidemiologic, Prevalence, Multivariate Analysis, Confidence Intervals, Logistic Models, Proportional Hazards Models, Case-Control Studies, Sample Size
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.